Bull. Mater. Sci., Vol. 37, No. 6, October 2014, pp. 1439–1445. © Indian Academy of Sciences.

Statistical analysis of thermal conductivity of nanofluid containing

decorated multi-walled carbon nanotubes with TiO2 nanoparticles

SEDIGHEH ABBASI, SEYED MOJTABA ZEBARJAD*,†, SEYED HOSSEIN NOIE BAGHBAN

and ABBAS YOUSSEFI††
Department of Chemical Engineering, Faculty of Engineering, Ferdowsi University of Mashhad, Azadi Square,
Mashhad, Iran
†
Department of Material Science and Engineering, Faculty of Engineering, Ferdowsi University of Mashhad,
Azadi Square, Mashhad, Iran
††
Par-e-Tavous Research Institute, Mashhad, Iran

MS received 16 June 2013; revised 1 September 2013

Abstract. In this paper, we report for the first time the statistical analysis of thermal conductivity of
nanofluids containing TiO2 nanoparticles, pristine MWCNTs and decorated MWCNTs with different amounts of
TiO2 nanoparticles. The functionalized MWCNT and synthesized hybrid of MWCNT–TiO2 were character-
ized using transmission electron microscopy (TEM). TEM image confirmed that the ends of MWCNTs were
opened during their oxidation of them in HNO3 and TiO2 nanoparticles successfully attach to the outer surface of
oxidized MWCNTs. Thermal conductivity measurements of nanofluids were analysed via two-factor completely
randomized design and comparison of data means is carried out with Duncan’s multiple-range test. Statistical
analysis of experimental data show that temperature and weight fraction have a reasonable impact on the
thermal conductivity of all tested nanofluids (α = 0⋅05). The results also show that increased temperature and
weight fraction leads to the increased thermal conductivity.

Keywords. MWCNT; TiO2 nanoparticles; thermal conductivity; nanofluid.

1. Introduction reported that the surfactant molecules which introduce

to the outer surface of nanoparticles and CNTs
Dispersing the particles with a typical size of < 100 nm in
increased the thermal resistance. Thus the application of
the conventional heat transfer fluids such as oil and water
surfactant decreased the thermal conductivity of nano-
leads to the production of new classes of heat transfer
fluids (Huxtable et al 2003). Introducing the oxygen-
fluids, which are called nanofluids (Arani and Amani
containing groups such as hydroxyl and carboxyl groups
2012). The application of particles with nanometer size
to the outer surface of carbon nanotubes has a significant
reduced the clogging of flow channels, eroding of pipe-
effect on the high stability of CNTs nanofluids, which leads
lines and pressure drop. In addition, compared with
to the enhancement of thermal conductivity (Kyotani et al
millimetre- or micrometer-sized particle suspensions, nano-
2001).
fluid have better long-term stability and greater thermal
Murshed et al (2005) reported the thermal conductivity
conductivity. Up to date, various types of nanoparticles
of TiO2/water-based nanofluids as a function of the shape
such as metallic, non-metallic, polymeric and nanotubes
of nanoparticles. They investigated TiO2 nanoparticles of
were suspended in base fluids (Duangthongsuk and
rod shapes of 10 nm × 40 nm (diameter by length) and in
Wongwises 2009). Among the various kinds of nano-
sphere shapes of 15 nm. Their experimental results show
particles, carbon nanotubes (CNTs), owing to their unique
that the thermal conductivity of nanofluids was much
structure, high aspect ratio and excellent thermal conduc-
higher than that of base fluid even in a small amount of
tivity, have attracted much attention. As a result, dispers-
nanoparticles. Meanwhile, they observed that volume
ing CNTs in a base fluid can enhance the thermal
fraction, particle size and shape of nanoparticles also
conductivity of nanofluid (Phuoc et al 2011). Dispersibi-
influence the thermal conductivity enhancement of nano-
lity of nanoparticles in the base fluid is a key factor in the
fluids.
stability of nanofluids. Therefore, the application of dis-
Zhang et al (2007) investigated the heat transfer per-
persing agent such as surfactant can enhance the stability
formance of TiO2/water nanofluid for various volume
of nanofluids (Assael et al 2005). The previous study
fractions and temperatures. They observed that the effec-
tive thermal conductivities of nanofluids have not shown
*Author for correspondence (zebarjad@um.ac.ir) any anomalous enhancements.
1439
1440 Sedigheh Abbasi et al

Chen et al (2008) reported the thermal conductivity of conductivity, such that we could investigate whether the
nanofluids containing multi-walled carbon nanotubes thermal conductivity enhancement of nanofluids contain-
(MWCNTs), which were treated using mechanochemical ing the hybrid of TiO2 nanoparticles and MWCNTs are
reaction method to enhance their dispersibility. Their considerable or not.
results showed the thermal conductivity enhancement up Till now, no paper has reported the effect of decorated
to 17⋅5% at volume fraction of 0⋅01. MWCNTs with metal oxide such as TiO2 nanoparticles
Raykar and Singh (2010) investigated the thermal con- with different amounts of nanoparticles loading. Mean-
ductivity of nanofluids containing carbon nanotubes as a while, the statistical analysis of the thermal conductivity
function of temperature and Brownian motion. They of these nanofluids has never been investigated before.
observed that Brownian motion played a key role in the Therefore, in this paper, we want to investigate, for the
thermal conductivity of nanofluids. first time, the statistical analysis of the thermal conduc-
Talaei et al (2011) studied the thermal conductivity of tivity of nanofluid containing modified MWCNTs with
nanofluids containing MWCNTs with various concentra- various amounts of TiO2 nanoparticles.
tions of functional group. They reported that the concen-
tration of functional groups which attached to the surface
of MWCNTs in different methods is varied. Meanwhile 2. Materials and experimental
they observed that thermal conductivity of nanofluids
increased with respect to the functionalized group con- TiO2 nanoparticles were prepared by the hydrolysis of
centration. titanium tetrachloride (TiCl4, M = 189⋅79, 99%, Merck).
Xie and Chen (2009) investigated the thermal The synthesis procedure of TiO2 nanoparticles is as
conductivity of nanofluids of MWCNs into ethylene gly- follows. A certain amount of TiCl4 was dissolved in
col base fluid. Their experiments showed that the thermal distilled water at ambient temperature and stirred for
conductivity was enhanced with nanotube loading and about 5 h. Then the temperature of the solution was
temperature. increased to 80 °C and stirred for 24 h. Finally, the solution
Meibodi et al (2010) have reported the thermal conduc- was filtrated using the Whatman filter paper (Ashless,
tivity and stability of CNT/water nanofluids as a function circies, 125 mm) and the separated TiO2 nanoparticles
of nanoparticle size and concentration, surfactant type dried at room temperature and calcined at 370 °C for 3 h.
and concentration, pH, temperature, power of ultrasonica- For the synthesis of decorated MWCNTs with TiO2
tion and elapsed time after ultrasonication and their inter- nanoparticles, the following materials are needed.
actions. Their results revealed that, as a result of the MWCNTs (average diameter of 40–60 nm and lengths
statistical analysis of the thermal conductivity, elapsed ranging from 5 to 15 μm), nitric acid (HNO3, M = 63,
time after ultrasonication had the most significant effect 65%, Merck), hydrochloride acid (HCl 37 wt%, Merck)
and pH was not more significant. and titanium tetrachloride which is mentioned earlier.
Jha and Ramaprabhu (2009) investigated the thermal The details of the synthesis procedure of MWCNT–TiO2
conductivity of nanofluids containing MWCNTs which has been described in our previous work (Abbasi et al
were decorated with Pd, Au and Ag crystalline metal 2013). Only a brief description is given here. The synthe-
nanoparticle. They used water and ethylene glycol as sis includes two steps. First, oxidation of MWCNTs with
base fluids without any surfactant. Their experiments nitric acid for about 2 h in an ultrasound bath and 2 h in a
corroborated that the thermal conductivity of these stirrer at high speed to open the end of MWCNTs and
nanofluids increases with increase in concentration of introduce the oxygen-containing groups to their outer
metal-MWNT and with temperature. surface. Then, the oxidized MWCNTs were washed with
Nevertheless the effect of many factors such as tem- distilled water until the pH value of the drained water
perature, pH, concentration, size, shape and type of reached neutral and dried overnight. Secondly, a certain
nanoparticles on the thermal conductivity of nanofluids amount of TiCl4 was dissolved in 100 mL distilled water,
have been reported by many researchers in the previous followed by adding a little HCl (37 wt%) to the distilled
studies. In addition, a few papers have been published on water before TiCl 4 was dissolved in the water. Then,
the thermal conductivity of nanofluids containing deco- ~ 75 mg of the functionalized MWCNTs were added to
rated CNTs with nanoparticles. Although the thermal the solution and dispersed using ultrasound bath for 2 h
conductivity enhancement in the nanofluid containing and stirred for 22 h. Finally, the temperature of solution
decorated MWCNT with Pd, Au and Ag crystalline metal increased to 80 °C and stirred for 3 h. then filtered, dried
nanoparticle was investigated by previous investigators, at 80 °C for 1 h and calcined at 370 °C for 3 h.
it is clear that these nanoparticles have high intrinsic Thermal conductivity of nanofluids was measured using
thermal conductivity. Therefore, the hybrid of these KD2 Prothermal property analyser purchased from Deca-
nanoparticles and MWCNTs can enhance the thermal gon Devices Inc. The measurements were carried out
conductivity of nanofluids. In this paper, we decided to using the single needle (KS-1) of 60 mm length and
choose TiO2 nanoparticles with low intrinsic thermal 1⋅3 mm diameter and accuracy ± 0⋅01 Wm–1 K–1 from
 Thermal conductivity of nanofluid containing decorated multi-walled carbon nanotubes 1441

0⋅02 to 0⋅2 Wm –1 K–1. The effect of temperature and con- of TiO2 nanoparticles by changing the amount of TiCl4.
centration was studied in the range of 25–70 °C and 0⋅25– Figure 2 illustrates TEM image of decorated MWCNTs.
1⋅5 wt%, respectively. For controlling the temperature dur-
ing the measurements, a water bath (Thermo Haake K10 3.3 Data analysis and statistical studies
TT4310) at constant temperature which circulated the
water around the double-walled cylindrical container of 3.3a Effect of temperature on thermal conductivity of
samples was connected to the KD2 Pro device. For accu- nanofluids: Figure 3 depicts the effect of temperature
rate results, the measurements were repeated several
times and only three measurements which had the mean
correlation coefficient r2 > 0⋅9998 were considered.
Transmission electron microscopy (TEM) micrographs of
acid-treated and modified MWCNTs with TiO2 nano-
particles were obtained using a LEO 912AB system
which operated at 120 kV.
Statistical analysis of the thermal conductivity of nano-
fluids in different nanofluids was performed with a two-
factor completely randomized design using Mstac (Ver.
1⋅42) and significant differences between means were
determined by Duncan’s multiple range test. P values
< 0⋅05 were considered statistically significant. For the
investigation, the combined effects of temperature and
weight fraction on the thermal conductivity of nanofluids,
response surface method was measured out using Minitab
Release software (Ver. 11.12).

3. Results and discussion

3.1 Oxidation of MWCNTs

As a result of oxidation of MWCNTs in nitric acid, the

Figure 1. TEM image of acid-treated MWCNTs.
ends of tubes which closed with catalyst particles were
opened. Furthermore, the acid treatment of MWCNTs
leads to the cutting of MWCNTs to a shorter length.
During oxidation of MWCNTs, the sp2 hybridization of
MWCNTs can be changed to sp3. Therefore, the outer
surface and open ends of MWCNTs functionalized
with oxygen-containing groups such as acid carboxylic
(–COOH) and hydroxyl (–OH). Figure 1 shows TEM
image of acid treatment of MWCNTs in nitric acid.

3.2 Decoration of MWCNTs

Hydrolysis of TiCl4 in the solution of functionalized

MWCNTs leads to the production of titanium ions. These
ions can be absorbed by hydroxyl and acid carboxylic
groups which introduce to the outer surface of MWCNTs
during the acid treatment process. These functional
groups act as active sites and titanium ions attach due to
electrostatic attraction. Therefore, the nucleation of TiO2
nanoparticles carried out and the calcination of the
powder leads to the formation of nanocrystalline TiO2 on
the outer surface of MWCNTs. It should be mentioned
that the amount of TiCl4 has a key role in the nucleation
and formation of TiO2 nanoparticles. Therefore, in this
study we synthesized MWCNTs–TiO2 with 34 and 61% Figure 2. TEM image of MWCNTs–TiO2.
1442 Sedigheh Abbasi et al

on the thermal conductivity of TiO2 nanofluids. As can be Figures 5 and 6 illustrate the effect of temperature on
deduced, there was a completely significant difference the thermal conductivity of nanofluids containing
among different temperatures from the aspect of thermal MWCNTs–TiO2 with 61 and 34% of TiO2 nanoparticles,
conductivity. By increasing the temperature from 25 to respectively. It is clear that in both the nanofluids, there
70 °C, the thermal conductivity of TiO2 nanofluid was a significant difference among four temperatures for
was increased. Index was increased from 25 °C thermal conductivity of MWCNTs–TiO2 nanofluid
(0⋅5907 Wm–1 K–1 ) to 40 °C (0⋅6426 Wm–1 K–1) equal to (α = 0⋅05) and the thermal conductivity of both nanofluids
8⋅78%; from 40 to 55° C (0⋅6623 Wm–1 K–1) equal to 3⋅06% increased with respect to the temperature. From figures 5
and from 55 to 70 °C (0⋅686 Wm–1 K–1) equal to 3⋅57%. and 6, it can be inferred that the least thermal conductivity
In addition, the thermal conductivity of TiO2 nanofluid was recorded for temperature of 25 °C equal to
during the increase of the temperature, from 25 to 70 °C, 0⋅6342 Wm–1 K–1 and 0⋅652 Wm–1 K–1 for nanofluids
was increased to ~ 16%. containing MWCNTs–TiO2 with 61 and 34% of TiO2
The influence of temperature on the thermal conducti- nanoparticles, respectively. In addition, maximum ther-
vity of pristine MWCNTs nanofluid is shown in figure 4. mal conductivity was related to the temperature of 70 °C
It is clear that there was a significant difference among
four temperatures for thermal conductivity of pristine
MWCNTs nanofluid (α = 0⋅05). Moreover, it can be
observed that by increasing the temperature, thermal con-
ductivity also increased. Minimum thermal conductivity
was related to a temperature of 25 °C equal to 0⋅6181
Wm–1 K–1. The value of this quality at 40 °C was increased
to 0⋅6852 Wm–1 K–1, e.g. 10⋅85%, at 55 °C increased to
0⋅7202 Wm–1 K–1, e.g. 5⋅10% and at 70 °C increased
to 0⋅7572 Wm–1 K–1, e.g. 5⋅13%. Therefore, the augmen-
tation of the thermal conductivity of pristine MWCNTs
nanofluid from 25 to 70 °C is equal to 22⋅50%. Mean-
while, comparing the augmentation of the thermal
conductivity of pristine MWCNTs nanofluid and
TiO2 nanofluid revealed that increasing thermal conduc-
tivity of nanofluid containing pristine MWCNTs nano-
fluid (22⋅5%) is > TiO2 nanofluid (16%). This can be due
to the excellent intrinsic thermal conductivity of
MWCNTs. In other words, the augmentation of the Figure 3. Effect of temperature on thermal conductivity of
thermal conductivity of pristine MWCNTs nanofluid is TiO2 nanofluid, means with different letters are significantly
40⋅62% greater than that of TiO2 nanofluid. Because in different based on Duncan’s multiple range test (α = 0⋅05).
the nanofluids containing sphere metal or metal oxide
nanoparticles, the Brownian motion of nanoparticles
affected the variation of thermal conductivity with tempe-
rature (Das et al 2003). Nevertheless, in the nanofluids
containing MWCNTs, in addition to the Brownian motion
of MWCNTs, the chemical functionalized groups have a
key effect on the amount of energy which transfers into
the nanofluids by changing the temperature (Chen and
Xie 2010).
It is clear that an increase in the temperature leads to
the weakness of the hydrogen bond of water molecules.
As a result, the number of free water molecules in the
nanofluid increases. Therefore, the produced free water
molecules arrange around MWCNTs’ surface as a liquid
layer due to the chemical surfaces of MWCNTs and van
der Waals force between the water molecules. The ther-
mal conductivity of the produced layer is higher than that
of bulk liquid. Therefore, it can be deduced that the
Figure 4. Effect of temperature on thermal conductivity of
chemical functionalized groups of MWCNTs have a MWCNT nanofluid, means with different letters are signifi-
significant effect on the enhancement of thermal conduc- cantly different based on Duncan’s multiple range test
tivity of nanofluid. (α = 0⋅05).
 Thermal conductivity of nanofluid containing decorated multi-walled carbon nanotubes 1443

equal to 0⋅7737 Wm–1 K–1 and 0⋅7896 Wm–1 K–1 for fractions in statistical level of 5% on aspect of thermal
nanofluids containing MWCNTs–TiO2 with 61 and 34% conductivity of TiO2 nanofluids. Maximum thermal
of TiO2 nanoparticles, respectively. conductivity of TiO2 nanofluids was 0⋅6512 Wm –1 K–1 at
Meanwhile, it can be deduced that in all the tested weight fraction of 1⋅5 wt% but its difference in compari-
temperatures, by increasing the loading of TiO2 nanopar- son with thermal conductivity at weight fraction of 1 wt%
ticles on the surface of MWCNTs, the thermal conducti- (0⋅6475 Wm–1 K–1 ) and 0⋅5 wt% (0⋅6437 Wm–1 K–1) was
vity of MWCNT–TiO2 nanofluids also increased. It is due insignificant. Minimum thermal conductivity was related
to the decreasing TiO2 nanoparticles content in the hybrid to the weight fraction of 0⋅25 wt% (0⋅6393 Wm–1 K–1 )
of MWCNTs–TiO2, the amount of MWCNTs which have that it was 1⋅18% less than that of weight fraction of
higher intrinsic thermal conductivity increased. 1⋅5 wt% (figure 7).
There was a significant difference among different
3.3b Effect of concentration on thermal conductivity of weight fractions from the viewpoint of thermal conducti-
nanofluids: The results of data variance showed that vity of MWCNT nanofluids. It was observed that by in-
there was a significant difference among four weight creasing the weight fraction of MWCNT, thermal

Figure 5. Effect of temperature on thermal conductivity of

MWCNT–TiO2 (61%) nanofluid, means with different letters Figure 7. Effect of weight fraction on thermal conductivity of
are significantly different based on Duncan’s multiple range TiO2 nanofluid, means with different letters are significantly
test (α = 0⋅05). different based on Duncan’s multiple range test (α = 0⋅05).

Figure 6. Effect of temperature on thermal conductivity of Figure 8. Effect of weight fraction on thermal conductivity of
MWCNT–TiO2 (34%) nanofluid, means with different letters MWCNT nanofluid, means with different letters are signifi-
are significantly different based on Duncan’s multiple range cantly different based on Duncan’s multiple range test
test (α = 0⋅05). (α = 0⋅05).
1444 Sedigheh Abbasi et al

conductivity also increased. Maximum thermal conduc- The influence of weight fraction on the thermal con-
tivity of MWCNT nanofluids was 0⋅7063 Wm–1 K–1 at ductivity of nanofluids containing MWCNT–TiO2 is
weight fraction of 1⋅5 wt% but its difference in compari- shown in figures 9 and 10 for MWCNT–TiO2 with 61 and
son with thermal conductivity at weight fraction of 1 wt% 34% TiO2 nanoparticles, respectively. It can be inferred
(0⋅699 Wm –1 K–1) was insignificant. Meanwhile, the that the minimum and maximum of thermal conductivity
thermal conductivity of MWCNT nanofluids during the of both MWCNT–TiO2 nanofluids were related to weight
increasing the weight fraction, from 0⋅5 to 1⋅5 wt% was fraction of 0⋅25 and 1⋅5 wt%, respectively. According to
increased equal to 3⋅36% (figure 8). Comparing the aug- the Duncan’s multiple range test (α = 0⋅05), there was a
mentation of the thermal conductivity of pristine MWCNTs significant difference among the highest and lowest
nanofluid with temperature and concentration revealed weight fractions for thermal conductivity of both
that the increasing of thermal conductivity of nanofluid MWCNT–TiO2 nanofluids, but difference between the
with temperature (22⋅5%) is higher than that of with weight fraction of 0⋅5 and 1 wt% was insignificant. In
concentration (3⋅36%), which is consistent with the addition, increasing the weight fraction from 0⋅25 to
results obtained by Ding et al (2006). 1⋅5 wt% led to the augmentation of thermal conductivity
equals to 2⋅975 and 3⋅416% for MWCNT–TiO2 nano-
fluids with 61 and 34% TiO2 nanoparticles, respectively.
Therefore, it can be deduced that increasing the thermal
conductivity of MWCNT–TiO2 nanofluids with 34%
TiO2 nanoparticles is higher than that of MWCNT–TiO2
nanofluids with 61% TiO2 nanoparticles. It is due to the

Figure 9. Effect of weight fraction on thermal conductivity of

MWCNT–TiO2 (61%) nanofluid, means with different letters
are significantly different based on Duncan’s multiple range
test (α = 0⋅05).

Figure 11. Response surface of thermal conductivity of TiO2

nanofluid vs temperature and weight fraction.

Figure 10. Effect of weight fraction on thermal conductivity

of MWCNT–TiO2 (34%) nanofluid, means with different letters
are significantly different based on Duncan’s multiple range Figure 12. Response surface of thermal conductivity of
test (α = 0⋅05). MWCNT nanofluid vs temperature and weight fraction.
 Thermal conductivity of nanofluid containing decorated multi-walled carbon nanotubes 1445

cases, it is observed that thermal conductivity value

increases with temperature and weight fraction. However,
the influence of temperature is more considerable than
that of weight fraction.

4. Conclusions

The present study investigated the thermal conductivity

nanotubes containing TiO2 nanoparticle, pristine and deco-
rated multi-walled carbon with TiO2 nanoparticles nano-
tubes. Statistical analysis of the experimental data shows
that the temperature and weight fractions have a significant
effect on the thermal conductivity of nanofluids. How-
Figure 13. Response surface of thermal conductivity of ever, the influence of temperature on the thermal conduc-
MWCNT–TiO2 (61%) nanofluid vs temperature and weight tivity of nanofluids is more reasonable than that of weight
fraction. fraction. Meanwhile, the results revealed that in all the
tested temperatures, by increasing the loading of TiO2
nanoparticles on the surface of MWCNTs, the thermal
conductivity of MWCNT–TiO2 nanofluids also increased.

Acknowledgements

The authors gratefully acknowledge the help by Mrs

R Pesyan from Central Research Laboratory of Ferdowsi
University of Mashhad and Mrs T Saeedi from Par-e-
Tavous Research Institute.

References

Abbasi S, Zebarjad S M and Baghban S H N 2013 Engineerng 5

207
Figure 14. Response surface of thermal conductivity of Arani A A A and Amani J 2012 Exp. Therm. Fluid Sci. 42 107
MWCNT–TiO2 (34%) nanofluid vs temperature and weight Assael M J, Metaxa I N, Arvanitidis J, Christofilos D and
fraction. Lioutas C 2005 Int. J. Thermophys. 26 647
Chen L, Xie H, Li Y and Yu W 2008 Thermochim. Acta 477 21
increasing portion of MWCNT with higher intrinsic Chen L and Xie H 2010 Thermochim Acta 497 67
thermal conductivity in the hybrid of MWCNT–TiO2. Das S K, Putra N, Thiesen P and Roetzel W 2003 J. Heat
Transfer. 125 567
Ding Y, Alias H, Wen D and Williams R A 2006 Int. J. Heat
3.3c Response surface study: In this study, the response
Mass Trans. 49 240
surface fitting method was used in order to study the Duangthongsuk W and Wongwises S 2009 Int. J. Heat Mass
combined effect of temperature and weight fraction on Tran. 52 2059
the thermal conductivity of nanofluid. Figure 11 shows Huxtable S T et al 2003 Nature Mater. 2 731
the surface response of thermal conductivity of TiO2 Jha N and Ramaprabhu S 2009 J. Appl. Phys. 084317 106
nanofluid, which implies that the thermal conductivity of Kyotani T, Nakazaki S, Xu W and Tomita A 2001 Carbon 39
TiO2 nanofluid increases with increasing temperature and 771
weight fraction, although the weight fraction variations of Meibodi M E, Sefti M V, Rashidi A M, Amrollahi A, Tabasi M
the thermal conductivity is less than that of temperature. and Kalal H S 2010 Int Commun. Heat Mass. 37 319
With respect to figure 12, which shows the surface Murshed S M S, Leong K C and Yang C 2005 Int. J. Therm.
response of thermal conductivity of MWCNT nanofluid Sci. 44 367
Phuoc T X, Massoudi M and Chen R H 2011 Int. J. Therm. Sci.
with temperature and weight fraction, thermal conducti-
50 12
vity increases with increasing temperature and weight Raykar V S and Singh A K 2010 Phys. Lett. A374 4618
fraction. Talaei Z, Mahjoub A R, Rashidi A M, Amrollahi A and
Figures 13 and 14 show the surface response of the Meibodi M E 2011 Int. Commun. Heat Mass. 38 513
thermal conductivity of MWCNT–TiO2 nanofluids with Xie H and Chen L 2009 Phys. Lett. A373 1861
61 and 34% TiO2 nanoparticles, respectively. In both the Zhang X, Gu H and Fujii M 2007 Exp. Therm. Fluid Sci. 31 593